Selection of starter cultures for the production of sour cassava starch in a pilot-scale fermentation process

h tt p : / / w w w . b j m i c r o b i o l . c o m . b r /

Biotechnology

and

Industrial

Microbiology

Selection

of

starter

cultures

for

the

production

of

sour

cassava

starch

in

a

pilot-scale

fermentation

process

Fernanda

Corrêa

Leal

Penido

_,

_Fernanda

_Barbosa

_Piló

_,

Sávio

Henrique

de

Cicco

Sandes

_,

_Álvaro

_Cantini

_Nunes

_,

_Gecernir

_Colen

_,

Evelyn

de

Souza

Oliveira

_,

_Carlos

_Augusto

_Rosa

_,

_Inayara

_Cristina

_Alves

_Lacerda

a_{UniversidadeFederaldeMinasGerais,FaculdadedeFarmácia,DepartamentodeAlimentos,BeloHorizonte,MG,Brazil}

b_{UniversidadeFederaldeMinasGerais,InstitutodeCiênciasBiológicas,DepartamentodeMicrobiologia,BeloHorizonte,MG,Brazil} c_{UniversidadeFederaldeMinasGerais,InstitutodeCiênciasBiológicas,DepartamentodeBiologiaGeral,BeloHorizonte,MG,Brazil}

a

r

t

i

c

l

e

i

n

f

o

Received18May2017 Accepted5February2018 Availableonline28February2018 AssociateEditor:SolangeI. Mussatto

Keywords:

Lacticacidbacteria Yeasts

Startercultures Fermentation Bakeryproducts

a

b

s

t

r

a

c

t

Sourcassavastarch(Polvilhoazedo)isobtainedfromaspontaneousfermentationconducted bymicroorganismsfromrawmaterialsandfermentationtanks.Thisproductis tradition-allyusedinthebakingindustryforthemanufactureofbiscuitsandBraziliancheesebreads. However,theendoffermentationisevaluatedempirically,andtheprocessoccurswithout standardization,whichresultsinproductsofinconsistentquality.Predominantmicrobiota fromacassavaflourmanufacturerwasisolatedinordertoselectstarterculturesforthe productionofsourcassavastarchinapilot-scalefermentationprocess.Lacticacid bacte-riaand yeasts wereisolated,enumerated andgroupedbyRestrictionFragment Length Polymorphism,andPCRfingerprinting,respectively.Oneisolateofeachmolecularprofile wasidentifiedbysequencingoftherRNAgene.LABwereprevalentthroughouttheentire process.Lactobacillusbrevis(21.5%),whichproducedthehighestvaluesofacidity,and Lac-tobacillusplantarum(13.9%)wereamongthemostfrequentspecies.Pichiascutulata(52.2%) wastheprevalentyeastandshowedamylolyticactivity.Theaforementionedspecieswere testedassingleandmixedstarterculturesinapilot-scalefermentationprocessfor28days.

L.plantarumexhibitedbetterperformanceasastarterculture,whichsuggestsitspotential fortheproductionofsourcassavastarch.

©2018SociedadeBrasileiradeMicrobiologia.PublishedbyElsevierEditoraLtda.Thisis anopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/

licenses/by-nc-nd/4.0/).

∗ _{Correspondingauthor.}

E-mail:fclpenido@gmail.com(F.C.Penido).

https://doi.org/10.1016/j.bjm.2018.02.001

1517-8382/©2018SociedadeBrasileiradeMicrobiologia.PublishedbyElsevierEditoraLtda.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction

Becausecassava(ManihotesculentaCRANTZ)hasahighstarch content (approximately 80%), it is an important source of carbohydratesthatcanbesoldfreshorprocessedintoa vari-ety ofvalue-added products; cassava ischaracterized as a multipurposecrop.1_{In2014,theestimatedBraziliancassava}

productionwas23milliontons.2_{Althoughwidelyconsumed,}

cassavahaslimitationsduetoitsperishability,toxicityand lowproteincontent.3_{Therootofcassavacontainscyanogenic}

glycosidesthatactasdefencesubstancesthroughtherelease ofhydrogencyanide,whichisresponsibleforitstoxicity.The traditional fermentations of cassava are quite suitable for thedetoxification,preservationanddevelopmentofproducts withdesirableviscoelastictexture.4_{Lacticfermentationnot}

onlyextendstheshelflifeofthisrootbutalsodecreasesits toxicity.5

Naturalfermentationofwetstarchextractedfromcassava rootisatraditionaltechnologywidelyusedinLatin Amer-ica.Sweet cassava starchand sour cassavastarch undergo thesameprocessofstarchextraction,butdifferin fermen-tationtime(fromtwotosevendaysandfrom20to70days, respectively) and therefore have different levels of acidity, maximum of1% ofacidity and 5% ofacidity, respectively. Sourcassavastarchstandsoutfromsweetcassavastarchand otherflourinthepreparationofbakeryproductsforitsunique expansioncapacity,withouttheadditionofbakingsodaand in the absence of gluten. Sour cassava starch is a typical Brazilianfoodgenerallyproducedbysmallandmedium-sized ruralindustries.6_{Assourcassavastarchisessentially}

hand-crafted,eventhoughitisproducedinmoderncassavaflour manufacturers, it still has heterogeneous physicochemical andsensoryquality.Thesuccessionofmicroorganismsfrom rawmaterialsandfermentationtanksoccursnaturally dur-ingcassavafermentationandresultsinamicrobiotawitha prevalenceoflacticacidbacteria(LAB),suchasLactobacillus plantarum,Lactobacillusfermentum,Lactobacillusbrevisand Leu-conostocmesenteroides.7–10 _{Inthe productionofsourcassava}

starchinBrazil,Lactobacillusoccursinassociationwithyeasts, suchasGalactomycesgeothricumandIssatchenkia(nowPichia)

sp.6,11

Thewiderange andcomplexity ofcassavaspontaneous fermentationmicrobiotaarethemainfactorsresponsiblefor thelackofhomogeneityandlowproductquality.Theuseof selectedstrainsisanimportantalternativebecauseitprovides lessvariationinthecontentofchemicalcompounds,shorter fermentationtime,higheryieldandsensorialquality.12,13_The

LAB,whichareusedasnaturalorselectedstarterculturesin fermentedfoods,areabletoacidifyandenhancetheflavor. Furthermore, the LAB can protect food from the develop-ment of pathogens due to the formation of antimicrobial compounds.14,15

AlthoughsourcassavastarchiswidelyconsumedinLatin Americaandgained prominenceinthe preparationof bak-eryproductswithalowlevelortheabsenceofgluten,there are fewstudieson itsfermentationprocess.Therefore, the studyofstarterculturescontributessignificantlytothe under-standingandoptimizationofsourcassavastarchproduction. Thepresentworkaimedtoselectthestartercultureswiththe

appropriatecharacteristicsfortheproductionofsourcassava starchinapilot-scalefermentationprocess.

Materials

and

methods

Collectionofsamplesfromcassavaflourmanufacturer

Atotalof16samplesof100geachwerecollectedfroma cas-savaflourmanufacturerinthemunicipalityofFormiga,Minas Gerais(MG)state,Brazil,ondays0,5,12,19,26,33and40of a56-dayspontaneousfermentationfromafermentationtank withusablecapacityof16,000L.Thesesampleswere trans-portedtothelaboratoriesofFoodMicrobiologyandIndustrial Microbiology and Biocatalysis(Faculdadede Farmácia, Uni-versidade Federal de Minas Gerais, MG,Brazil) on ice and processedwithin24h.Theprocessingconsistedofweighing twenty-fivegramsofeachsampleinsterileflasks,dilutedin 225mLof1gL−1peptonewater,andpreparingserialdecimal dilutions.16

Identificationoflacticacidbacteriafoundonsamplesfrom cassavaflourmanufacturer

AppropriatedecimaldilutionswerespreadondeMan,Rogosa andSharpe(MRS;Acumedia,Lansing,MI,USA)agar contain-ing0.1gL−1cycloheximide16_{andincubatedinanaerobicjars}

of 2.5L (Permution, Curitiba, Brazil)at 37◦C for 48h. After growth,onecolonyofeachmorphotypewascountedand puri-fiedforlateridentification.EachisolatewasGramstainedand thensubjectedtothecatalasetest.17

DNA was extracted with an adaptation of the method described by Hoffman and Winston.18 _{The colonies}

previ-ously grown on MRS agar were resuspended in 100␮L of Tris–EDTA (TE).Then, 100␮Lofphenol–chloroform–isoamyl alcohol(25:24:1) and0.3gofglassbeadswere addedtothe suspension. Tubes containing this mixture were homoge-nized by a vortex shaker (QL-901, Biomixer, Santa Clara, CA, USA) for three to4min and centrifuged at 13,000rpm for 5min(Eppendorf, Hamburg, Hamburg, Germany).After that,thesupernatantwastransferredtoanothertube.Then, a volume of 960mLL−1 ethanol corresponding to the vol-ume ofthe supernatant recovered was added to the tube. Thetubes were homogenizedbyinversionand centrifuged at 13,000rpm for 2min. The liquid phase was discarded, the tubes were driedovernight,and the DNA resuspended in 50␮LofTE. TheDNA concentration wasdetermined by NanoDropND1000Spectrophotometer(NanoDropProducts, Wilmington, DE,USA). TheDNA oflactic acid bacteriawas subjectedtoPCRamplificationofthe16SrRNAgeneusingthe primers27F(5-AGAGTTTGATCCTGGCTCAG-3)and1492R(5 -GGTTACCTTGTTACGACTT-3).19_{AllLABisolatesweregrouped}

byRestrictionFragmentLengthPolymorphism(RFLP)by diges-tionwithrestrictionenzymesMspI,HinfIandHaeIII(Promega Corporation, Madison, WI, USA) according to the modified methodology of Brightwell et al.20 _{For the digestion}

reac-tion,2␮Lof10×buffer,2␮Lofbovineserumalbumin(BSA) onlyfortheMspIenzyme,1␮Lofenzyme,DNA≤1500ng/␮L and water q.s.p. 20␮L. Thetubes were incubated at 37◦C for 3h. Therestriction fragments obtainedwere separated

by2%agarosegelelectrophoresis(Pronadisa,Spain)in0.5% TBE buffer at 100V. Thegels were stained withGelRedTM solution(Biotium,USA)andvisualizedunderultravioletlight (UV)byanimagecapturesystem(VilberLourmat,France).An isolateofeachdifferentmolecularRFLPprofilewasselected andsubjectedto16SrRNAsequenceanalysis.19_Sampleswere

sequencedbycapillaryelectrophoresisinABI3130equipment usingPOP7polymerandBigDyev3.1(MyleusBiotechnology, BeloHorizonte,MG,Brazil).Thesequenceswereassembled, editedandalignedwiththeprogramMEGA6.21_Thesequences

obtainedwerecomparedwiththoseincludedintheGenBank databaseusingtheBasicLocalAlignmentSearchTool(BLAST

athttp://www.ncbi.nlm.nih.gov).

DifferentiationofspeciesoftheL.plantarumgroup

Alltheisolatesthatwerepresumablyidentifiedby16SrRNA sequenceanalysisasbelongingtotheL.plantarumgroupwere subjected to Multiplex PCR Assay with recA Gene-Derived Primers. This analysis allows the separation of the three closelyrelatedspeciesfromtheL.plantarumgroupby com-parisonofthesizeoftheiramplicons:318bpforL.plantarum,

218bpforLactobacilluspentosus and 107bpforLactobacillus paraplantarum.22 _{L. plantarum ATCC1 4917, L. paraplantarum}

DSM10667andL.pentosusATCC8041wereusedascontrols.

Identificationofyeastsfoundonsamplesfromcassava flourmanufacturer

Portionsofappropriatedecimaldilutionswere spread onto yeastextract–maltextract(YM;Acumedia,Lansing,MI,USA) agarcontaining0.2gL−1chloramphenicolandincubatedfor 48hat25◦Cunderaerobicconditions.6 _{Onecolonyofeach}

differentmorphotypewascountedandpurifiedforlater iden-tification.

ThetotalDNA ofeach yeast isolatewas extracted with an adaptation of the method described by da Silva-Filho et al.23 The colonies previously grown on YM agar were resuspendedin100␮Loflysisbuffer(Tris–HCl– trishydrox-ymethylaminomethane–0.05M,0.05MEDTA,0.1MNaCland 10gL−1 SDS –sodium dodecyl sulfate). Thetubes contain-ingthesuspensionwereincubatedinawaterbathat65◦C for35min.Then,100␮Lofphenol–chloroform–isoamyl alco-hol(25:24:1)wereaddedtothesuspension.Tubescontaining thismixturewerehomogenizedbyavortexshaker(QL-901, Biomixer,SantaClara,CA,USA)for3minandcentrifugedfor 15min(Eppendorf,Hamburg,Hamburg,Germany).Afterthat, thesupernatant wastransferredto anothertube, towhich wasadded100␮Lofcold700mLL−1ethanol.Thetubeswere homogenizedbyinversionandcentrifugedat13,000rpmfor 3min(Eppendorf,Hamburg,Hamburg,Germany).Theliquid phasewasdiscardedandthetubesweredriedovernight.The DNAwasresuspendedin100␮LofTE,andstoredinafreezerat −20◦_{C.TheDNAconcentrationwasdeterminedbyNanoDrop}

ND1000Spectrophotometer(NanoDrop Products, Wilming-ton,DE,USA).

Allyeastisolatesweregroupedbytheirmolecularprofiles usingthePCRfingerprintingtechniquewiththemicrosatellite primer (GTG)5 (5-GTGGTGGTGGTGGTG-3).24 The PCR

mix-ture contained2.5␮Lof 10× buffer, 1␮Lof dNTPs0.1mM,

1.5␮LofMgCl2,2␮Loftheprimer(GTG)510␮mol−1

(Invitro-gen,Carlsbad,CA, USA),0.2␮LofTaqDNA polymerase1U and 1␮LofDNAina totalvolume of25␮L.ThePCR reac-tionwasperformedontheMastercycler®Proandshowedthe followingconditions:initialdenaturationat94◦C for2min, 40cyclesofdenaturationat95◦Cfor45s,annealingat50◦C for1minandextensionat72◦Cfor1min,followedbyfinal extensionat72◦Cfor6min.Onerepresentativeofeach dif-ferentmolecularprofilewasidentifiedbysequencingofthe D1/D2 domains of the large subunit of rRNA gene using theprimersNL1(5-GCATATCAATAAGCGGAGGAAAAG-3)and NL4(5-GGTCCGTGTTTCAAGACGG-3)accordingtoLachance etal.25_{ThePCRmixturecontained5␮Lof10×buffer,1␮Lof}

dNTPs0.05mM,3␮LofMgCl21.5mM,1␮LoftheprimerNL1

10␮mol−1_{(Invitrogen,Carlsbad,CA,USA),1␮Loftheprimer}

NL410␮mol−1(Invitrogen,Carlsbad,CA,USA),0.2␮LofTaq DNApolymerase1Uand1␮LofDNAinatotalvolumeof50␮L. Thereaction was performedon theMastercycler® Pro and showedthefollowingconditions:initialdenaturationat95◦C for2min,35cyclesofdenaturationat95◦Cfor15s,annealing at54◦Cfor25s,andextensionat72◦Cfor20s,followedbyfinal extensionat72◦Cfor10min.Thesamplesweresequenced bycapillaryelectrophoresisinABI3130equipmentusingPOP7 polymerandBigDyev3.1.Sequenceanalyseswereperformed asdescribedabove.

Evaluationofsamplesfromcassavaflourmanufacturer

Total titratable acidity (TTA) and pH of samples were determined.26_{Thefinalproductfromthefermentationtank}

was submitted to microbiological analyses to search for

Bacillus cereus,thermotolerantcoliforms andSalmonellaspp. accordingtoBrazilianlegislation.27

Screeningforselectionofstrainstobetestedassingleor mixedstarterculture

All thestrainsofLABspeciesisolatedwere testedin tripli-cateforstarchdegradationonplateswithMRSagarcontaining 20gL−1 solublestarch28 _{andallthe strainsofyeastspecies}

isolated,onplateswithYMagarcontaining20gL−1 soluble starch.29 _{After growth, the revelation was performed with}

iodinesolutioninordertoviewstarchhydrolysishalos.Total acid productionwas evaluated in100mLofbroth contain-ing20gL−1(non-fermented)cassavastarch,10gL−1 glucose and5gL−1ofbeefextractwith24and48h.Aliquotsof10mL ofbrothwereusedforTTAmeasurement.26_{Thestrainsthat}

showed starchdegradation halosandhighervaluesofTTA weresubsequentlytestedasstartercultures.

Useofstarterculturesinapilot-scalefermentationprocess

Single culturesofselected LAB strains and mixedcultures oftheseinassociationwiththe yeastwere inoculatedinto 100mLofculturemediumcontaining20gL−1(non-fermented) cassavastarch,10gL−1 glucoseand5gL−1 beefextractand thenincubatedatroomtemperaturefor24–48h.Theentire fermentedbrothwasusedtoinoculate500mLofthesame mediumfor24–48h.Theresultingbrothwas usedto inoc-ulate 5Lof a mediumcontaining 100gL−1 cassava starch,

for 28 days. The 5-L fermentations were inoculated with 8log10CFU/mLofLABand/or6log10CFU/mLofyeasts.Assays

wereperformedwithoutaeration.

Sampleswereweekly collectedintriplicatefor determi-nation of pH and TTA26 _{as well as for verification of the}

viabilityofcultures.Themonitoringofstartercultureswas donebyisolationandpurificationofthreecoloniesperweek ofeachLABoryeastthatresembledmorphologicallytothe starterculturesselectedandperformedwithmolecular meth-odsdescribedabove,andconductedbycomparisonofprofiles obtainedwithprofilesalready known.Afterthe endof fer-mentation,thematerialwasdriedatroomtemperaturefor 5days,packedandstoredunderrefrigerationuntil comple-tionofmeasurementsofpHandTTA26_{;expansioncapacity}30_;

andmicrobiologicalanalysis,asdescribedabove.The expan-sioncapacityassaywascarriedoutaccordingtotheprocedure proposed byTropicalRoots Center.Fortymilliliters of boil-ingdistilledwaterwereaddedto50gofcassavastarch.Each masswasmodeledformakingroundcookiesofapproximately 10geach.Thediametersofthecookiesweremeasuredwith auniversalpachymeter(Series530,Mitutoyo)beforeandafter takingthemtotheovenforaperiodof20minat220◦C.The expansioncapacitywas calculatedbythe ratio ofaverages ofthe finaldiameter(afterbaking)and theinitialdiameter (beforebaking).

Dataanalysis

The data obtained were subjected to analysis of variance (ANOVA) and significant differences between means were determinedbyTukey’stestwitha0.05significancelevel.

Results

Identificationofmicroorganismsfromcassavaflour manufacturer

LAB were present in all samples during a 56-day cassava fermentation with higher counts (5.8–7.9log10CFU/g) than

those obtained for yeasts (1.7–7.8log10CFU/g), which were

found from day 0 to 26. The 79 isolates of LAB identi-fiedwereGram-positive, catalasenegativeandmostly rods (83%).ThebacterialisolateswereefficientlygroupedbyRFLP

(Figs.1and 2)intwelvedifferent species.Sevenspeciesof

Lactobacilluswereisolated:L.brevis (n=17,21.5%),L. fermen-tum (n=12, 15.2%), L. plantarum (n=11, 13.9%), Lactobacillus casei/Lactobacillusparacasei(n=7,8.9%),Lactobacillusharbinensis

(n=5, 6.3%), Lactobacillus parabuchneri (n=4, 5.0%) and Lac-tobacillus ghanensis (n=1, 1.3%). Enterococcus faecium (n=13, 16.5%)wasalsoisolatedand,inminorproportions,Weisella cibaria(n=3,3.8%),Lactococcusgarvieae(n=3,3.8%),Lactococcus lactissubsp.lactis(n=2,2.5%)andL.mesenteroides(n=1,1.3%). ThemostfrequentisolateswereL.brevis,isolatedfromsixout ofsevensamplecollections(5.8–7.5log10CFU/g),andL.

plan-tarum,isolatedinfivesamplecollections(5.4–6.4log10CFU/g).

Alloftheelevenisolatesthatwerepresumablyidentified by16SrRNAsequenceanalysisasbelongingtotheL.plantarum

groupwereposteriorlyconfirmedasspeciesofL.plantarumby MultiplexPCRAssayusingrecAGene-DerivedPrimers.

The23yeastisolateswereefficientlygroupedbyanalysisof theirmolecularprofilesobtainedbyPCRfingerprinting(Fig.3). ThemostcommonlyisolatedyeastspecieswerePichia scutu-lata(n=12,52.2%)andKazachstaniaexigua(n=4,17.4%),both foundfromday0to12(>4log10CFU/g).Candidahumilis(n=3,

13.0%),Geotrichumfragrans(n=3,13.0%)andCandidaethanolica

(n=1,4.4%)werealsoisolated.

Evaluationofsamplesfromcassavaflourmanufacturer

The pH values decreased and there was a consequent increase in TTA (Table 1). Bacillus spp. were found in the finalproductsample,withcountsof9.33×102_CFU/g,which

is a valuelower than the tolerancepermitted byBrazilian law(3.00×103_CFU/g).31_{Neitherthermotolerantcoliformsnor}

Salmonellaspp.weredetected.

Screeningforselectionofstrainstobetestedassingleor mixedstarterculture

Noneofthe82LABstrainsisolatedexhibitedamylolytic activ-ity.However,outofthe33yeaststrainsisolatedalltheones belongingtothespeciesP.scutulata,C.humilisandC.ethanolica

showedpositiveresults(datanotshown)forstarch degrada-tion (indicatedbyaclearzone ofinhibition>1mmaround colonies).TTAvaluesevaluatedforthepredominant micro-biotaweregreaterat48hthanat24hoffermentationformost oftheLAB.ThebacteriawithhigheracidityvalueswereL. bre-visandL.garvieae.Tables2and3showtheresultsofTTAfor thebeststrainisolatedofeachspecies.

OneisolateofL.brevisandoneisolateofL.plantarumwere selectedtobetestedasthestarterculturesbecausetheywere amongthepredominantLABthathadhighervaluesofacidity. Theisolateofyeastchosentobetestedasstarterculturewas onestrainofP.scutulata,thepredominantyeastspecies,which exhibitedamylolyticactivity.Moreover,P.scutulataferments glucose,whichisimportantfortheproductionofsecondary metabolitessuchasvolatilecompoundsresponsibleforthe characteristicaromaofsour cassavastarch.In the present study,wetestedfourstartercultures:L.brevis;L.plantarum; L.brevisinamixedculturewithP.scutulata;andL.plantarum

inamixedculturewithP.scutulata.

Useofstarterculturesforsourcassavastarchprocessing

MonitoringofTTAandpHduringthepilot-scalefermentation processrevealedthatvaluesobtainedfordifferentstarter cul-turesdidnotdiffer(Fig.4).TTArangedfrom0.14%to0.71%. The pHrangedfrom 5.69to 3.29.ANOVA showed that the extentoffermentationsignificantlychangedtheTTAandpH ofcassavastarch.

The starter cultures used were able to prevail during cassava pilot-scale fermentation; therefore, all the isolates selectedforidentificationdidcorrespondtothestarter cul-turesinoculated.TheLABwasobservedtoremainin fermen-tationduring28days.ItwasobservedthatP.scutulatashowed amaximumcount,approximately7log10CFU/g,within7days

offermentation,andthegrowthdecreasedreachingcounts between3and4log10CFU/gwith14days.Thenumberofviable

10000 bp

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

1000 bp 500 bp 250 bp

Fig.1–Digestionprofileoflacticacidbacteriaisolatedfromcassavafermentationinacassavaflourmanufacturer.Lanes:1: 1kbDNAladder;2–4:Lactobacillusbrevisprofile(2-MspI,3-HaeIII,4-HinfI);5–7:Lactobacillusfermentumprofile(5-MspI, 6-HaeIII,7-HinfI);8–10:Lactobacillusplantarumprofile(8-MspI,9-HaeIII,10-HinfI);11–13:Lactobacilluscasei/L.paracaseiprofile (11-MspI,12-HaeIII,13-HinfI);14–16:Lactobacillusharbinensis(14-MspI,15-HaeIII,16-HinfI);17–19:Lactobacillusparabuchneri profile(17-MspI,18-HaeIII,19-HinfI).

10000 bp

500 bp 1000 bp

250 bp

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Fig.2–Digestionprofileoflacticacidbacteriaisolatedfromcassavafermentationinacassavaflourmanufacturer.Lanes:1: 1kbDNAladder;2–4:Lactobacillusghanensisprofile(2-MspI,3-HaeIII,4-HinfI);5–7:Enterococcusfaeciumprofile(5-MspI, 6-HaeIII,7-HinfI);8–10:Weisellacibariaprofile(8-MspI,9-HaeIII,10-HinfI);11–13:Lactococcusgarvieaeprofile(11-MspI, 12-HaeIII,13-HinfI);14–16:Lactobacilluslactisprofile(14-MspI,15-HaeIII,16-HinfI);17–19:Leuconostocmesenteroidesprofile (17-MspI,18-HaeIII,19-HinfI).

1 10000 bp 1500 bp 1000 bp 750 bp 2 3 4 5 6 7 8 9 10 11 12

Fig.3–PCRfingerprintingprofileofyeastsisolatedfrom cassavafermentationinacassavaflourmanufacturerin Formiga(MG,Brazil).Lanes:1:1kbDNAladder;2,5–7:

Kazachstaniaexigua;3,4and8:Candidahumilis;9and10:

Pichiascutulata;11:Candidaethanolica;12:Geotrichum

fragrans.

L.plantarumwheninassociationwiththeyeast.Onthe21st day,thepresenceoftheyeastwasnolongerobserved(Fig.5). Cassavastarch,thedriedfinalproduct,obtainedwithL. plantarumexhibitedthehighestTTAvalueamongthesamples.

Table1–TTAandpHofsamplesfromacassavaflour

manufacturerinthemunicipalityofFormiga,Minas

Gerais(MG)state,Brazil.

Time TTA pH

19 3.31 ±0.17 c _{4.01 ±0.01} a

26 6.83 ±0.25 b _{3.75 ±0.02} b

33 9.36 ±0.20 a _{3.53 ±0.07} cd

40 9.77 ±0.23 a _{3.46 ±0.02} d

Meanvalues±standarddeviationinthesamecolumnfollowedby differentsuperscriptlettersaresignificantlydifferent(p<0.001).

Althoughthisvaluewasnotclosetothevalueofcommercial product, it waswithinthe establishedlimits oflegislation,

whichmandates amaximumacidityof1.0%forsweet

cas-savastarchandamaximumacidityof5.0%forsourcassava starch.32_{Hence,theproductobtainedwithL.plantarum}

sin-gleculturecouldbecalledsourcassavastarch.Finalproducts obtainedwith other starter culturesmightbe calledsweet cassavastarch.ThepHvalueofstarchobtainedwithL. plan-tarum singleculturewas the onlyvaluethat didnotdiffer fromthepHvalueofthecommercialproduct.Theexpansion capacitiesforthecassavastarchesobtainedwiththestarter

Table2–TTAafter24and48hoffermentationwith

lacticacidbacteriaisolatedfromacassavaflour

manufacturerinthemunicipalityofFormiga,Minas

Gerais(MG)state,Brazil.

Species TTA(%)

24h 48h

Lb.brevis 3.47 ±0.06 ay _{5.28 ±0.07} ax

Lc.garvieae 3.37 ±0.20 ay _{5.28 ±0.14} ax

Lb.plantarum 2.96 ±0.08 ay _{4.97 ±0.19} abx

Lc.lactissubsp.lactis 1.63 ±0.09 ay _{4.47 ±0.24} abx

E.faecium 2.79 ±0.15 ay _{4.35 ±0.10} abx Ln.mesenteroides 2.74 ±0.44 ay _{3.77 ±0.03} abx Lb.parabuchneri 2.07 ±0.08 ay _{3.51 ±0.10} abx Lb.fermentum 2.57 ±0.09 ay _{3.50 ±0.03} abx W.cibaria 0.97 ±0.16 ay _{3.11 ±0.05} abx Lb.ghanensis 0.62 ±0.13 ay _{2.68 ±0.14} abx Lb.harbinensis 2.59 ±0.04 ax _{2.55 ±0.10} abx Lb.casei/Lb.paracasei 1.26 ±0.04 ay _{2.33 ±0.20} bx

Meanvalues±standarddeviationinthesamelinefollowedby dif-ferentsuperscriptletters(x,y)aresignificantlydifferent(p<0.001). Meanvalues±standarddeviationinthesamecolumnfollowedby differentsuperscriptletter(a,b)aresignificantlydifferent(p<0.001).

Table3–TTAafter24and48hoffermentationwith

yeastsisolatedfromacassavaflourmanufacturerinthe

municipalityofFormiga,MinasGerais(MG)state,Brazil.

Species TTA(%) 24h 48h Geotrichumfragrans 1.27 ±0.00 ay _{1.60 ±0.10} ax Pichiascutulata 0.48 ±0.10 ay _{0.78 ±0.05} ax Candidaethanolica 0.39 ±0.05 ax _{0.45 ±0.00} ax Candidahumilis 1.03 ±0.14 ax _{0.81 ±0.00} ay Kazachstaniaexigua 1.18 ±0.09 ax _{0.94 ±0.05} ay

Meanvalues±standarddeviationinthesamelinefollowedby dif-ferentsuperscriptletters(x,y)aresignificantlydifferent(p<0.001). Meanvalues±standarddeviationinthe samecolumn followed bydifferent superscriptletter(a) arenot significantly different (p>0.05).

culturesandforcommercialproductdidnotdiffer(Table4). Cassavastarchesproducedviaapilot-scalefermentation pro-cessprovedtobewithinthemicrobiologicallimits31_andwere

thereforesuitableforhumanconsumption.

Discussion

DuringtheisolationofLABandyeastsfromcassava fermen-tation,LABwerepresentinallsamplescollectedwithhigher counts than those obtained foryeasts. Many authors also reportedthepresenceofLABasprevalentmicroorganismsin cassavafermentations.33–35_{HighLABcountsareevidenceof}

theimportanceofthesemicroorganismsincassava fermen-tation.

AmongtheLABoftenfoundintraditionalcassava fermen-tations,manyauthorsreportthepresenceofL.brevisandL. plantarum,7,9,10,34,36,37 _{whichwere amongthemostfrequent}

isolatesinthepresentstudy.Lacerdaetal.6_reportedL.

plan-tarumandL.fermentumasthepredominantLABduringsour

cassavastarchfermentationisolatedfromtwocassavaflour manufacturerslocatedinthemunicipalityofConceic¸ãodos Ouros(MG,Brazil).L.breviswasisolatedinminorproportions. Thecountsofthesebacteriaweresimilartothosefoundin thisstudy.Themicrobiotafromcassavafermentationhas dif-ferentoriginsandmaycomefromrawmaterials,utensilsand equipmentusedinitsproduction.Insectsorhandlerscanalso carrythesemicroorganisms.

Themostcommonlyisolatedyeastspecies,P.scutulata,K. exigua,C.humilis,G.fragransandC.ethanolicawerealsofound inothercassavafermentationstudies.Lacerdaetal.6_collected

samplesfromsourcassavastarchfermentationandidentified thefollowingyeastspecies:Pichiasp.,genusthatwasformerly knownasIssatchenkiasp.,wasfoundincountsof5log10CFU/g;

C.humilisandC.ethanolicawerealsofrequentlyisolated. Wil-fridPadanou10_{foundaprevalenceofSaccharomycescerevisiae}

(22%)and P.scutulata(20%)inthe fermentationoflafun,an Africanproductobtainedfromcassavasubmerged fermenta-tion.Theoccurrenceoftheseyeastssuggeststhattheycould contributetothe sensorialquality. Severalstudies revealed thepresenceofthegenusCandidaincassavafermentationsto obtaindifferentproducts.6,7,10,38,39_{Inthepresentstudy,one}

speciesofthegenusPichiaandtwospeciesofthegenus Can-didawereisolated,showingthatspeciesofthesegeneracan growinacidicconditionsfoundincassavafermentationtanks. TheacidicenvironmentcreatedbyLABfavorsthe prolif-erationofyeastsinfoods.Simultaneously,thegrowthofLAB isstimulatedbythepresenceofyeasts,whichprovidegrowth factorssuchasvitaminsandnitrogencompounds.40_The

asso-ciationofLABandyeastsduringfermentationhasasignificant impactonfoodqualityparameterssuchastexture,flavorand nutritional values.41 _{Even though the complex interactions}

between LABand yeastsare not yetfully understood,it is knownthatLABandyeastshavetheabilitytoadapttomany differentsubstrates.42

Thedecrease ofpHvaluesand theconsequent increase in TTA of samples from cassava flour manufacturer were expectedresultsbecause,duringfermentation,thesynthesis oforganicacidsoccurs,leadingtoacidificationofthemedium. Itispossiblethatyeastshaveasmallcontributionordonot contributetoacidification,especiallybecausemostyeastsare not resistant to extreme acidic conditions. The pH values foundinthepresentstudyarelowerthanthosedetermined byCoulinetal.7_{inattiéké,afermentedcassavaproduct.The}

averagepHofthisfoodduringthetraditionalfermentationin smallscalewas5.0.

According to the microbiological analyses, sour cassava starch from the fermentation tank studied would be con-sideredsuitableforhumanconsumption.Verificationofthe presenceofB.cereusisimportantbecauseitproducestoxins andmaycausefoodpoisoningwhenconsumedinnumbers higherthan105_{CFU/g.Besidestheabilitytoproduceorganic}

acids,LABalsocontributetoinhibitionofpathogensdueto productionofantimicrobialcompoundssuchashydrogen per-oxide,diacetylandbacteriocins.43,44LABacidificationpoweris knowntobeeffectiveincontrollinggrowthofmicroorganisms infoodandextendingitsshelflife.45

The microorganisms considered as predominant in the presentstudyweretheoneswithhighercountsandthemost frequentlyisolatedduringcassavafermentationprocess.The

1,20 _6,00 5,00 4,00 3,00 2,00 1,00 0,00 1,00 0,80 0,60 0,40 0,20 0,00 0 7

Time (days) Time (days)

TT

A (%) pH

14 21 28 0 7 14 21 28

Fig.4–MonitoringofTTAandpHduringcassavapilot-scalefermentationwithselectedstartercultures.

8 7 6 5 4 3 2 1 0 0 7 14

Time (days) Time (days)

Time (days) Time (days)

Count (log 10 CFU/g) 8 7 6 5 4 3 2 1 0 Count (log 10 CFU/g) 8 7 6 5 4 3 2 1 0 Count (log 10 CFU/g) 8 7 6 5 4 3 2 1 0 Count (log 10 CFU/g) 21 28 0 7 14 21 28 0 7 14 21 28 0 7 14 21 28

Fig.5–Counts(log10CFU/g)oftheselectedstarterculturesduringcassavapilot-scalefermentation.

Table4–TTA,pHandexpansioncapacityforcassavastarches(driedfinalproducts)obtainedinapilotscale

fermentationcomparedwithacommercialproduct.

Cassavastarches TTA(%) pH Expansioncapacity

Lactobacillusbrevis 0.57 ±0.06 cd _{6.07 ±0.04} a _{1.04 ±0.00} a

Lactobacillusplantarum 1.41 ±0.00 b _{3.64 ±0.03} d _{0.98 ±0.04} a

L.brevis+Pichiascutulata 0.71 ±0.10 c _{4.34 ±0.06} c _{1.10 ±0.08} a

L.plantarum+P.scutulata 0.40 ±0.00 d _{4.84 ±0.02} b _{1.05 ±0.02} a

Commercialproduct 6.75 ±0.20 a _{3.62 ±0.04} d _{1.06 ±0.05} a

Meanvalues±standarddeviationinthesamecolumnfollowedbydifferentsuperscriptlettersaresignificantlydifferent(p<0.001).

resultsare inagreementwith the findingsofpredominant

microorganismsintraditionalcassavafermentations,

includ-ing sourcassava starchfermentation, asreported inmany

previousstudies.6–11,33–38,46–48_{Therefore,themicroorganisms}

isolated and identified inthe present study are a suitable

sourceofmicroorganismsforuseintestsfortheselectionof startercultures.

Selection of starter cultures showed that some yeast species exhibitedamylolytic activity. Thisobservation sug-gests that yeasts possibly play an important role at the

beginning of fermentation, degrading starch and releasing sugarsforthe growthofLABandtheir own.Ialsoshowed that TTA values obtained by LAB were higher than those obtainedbyyeasts,whichindicatesthattheformeraremainly responsible for acidification, which was also confirmed by OguntoyinboandDodd.35

L.plantarum,selectedasastarterculturefortheproduction ofsourcassavastarchinpilot-scalefermentationprocess,has alreadybeen evaluatedasastartercultureforthe fermen-tationofothertraditionalcassavaproduct.Kostineketal.33

assessedthe technological properties ofthe prevalent LAB inthefermentationofgari,anAfrican productobtainedby cassavasolid-statefermentation,andnotedthatL.plantarum

exhibitedbetterand fasteracid productionamongthe iso-latedbacteriaandwasrecommendedasthestarterculture. Huchetal.49_{testedL.plantarumandconcludedthatitssuccess}

to predominatein cassava fermentation demonstrates the potentialforitsdevelopmentasastartercultureforgari indus-trialization.Edwardetal.50_{investigatedtheuseoflyophilized}

LABstrainsasstarterculturesforgariproductionandfound thatL.plantarumcouldbeproducedatlowcost.

MonitoringofTTArevealed valueshigherthanthe ones determinedbyVogelmannetal.42_{whenstudyingcassava}

fer-mentedbyanassociationofLABandyeastsusedasstarter cultures.Thevaluesfoundbytheauthorsrangedfrom10.3to 11.5SH(DegreesSoxhlet-Henkel),whichcorrespondsto0.23% and0.26%,respectively.MonitoringofpHrevealedvalues simi-lartotheonesdeterminedbyVogelmannetal.,42whichvaried from4.1to4.2duringthe12daysoffermentation.

Themonitoringoftheviabilityofstarterculturesshowed that the culturesused were able toprevail duringcassava pilot-scale fermentation. This result may suggest a strong associationbetweenLABandyeastsatthebeginningof fer-mentation and indicates that LAB play an important role duringthe entireprocess,especiallyconcerningthequality andsafetyofthefinalproduct.Besidesthat,theuseofstarter culturesmay optimizethefermentationprocessdecreasing thefermentationtime.Huchetal.49_{testedL.plantarumand}

L.fermentumasstarterculturesfortheproductionofgari.The molecularmonitoringbyRandomAmplifiedPolymorphicDNA (RAPD)-PCRand typingtechniquesbyPulsedFieldGel Elec-trophoresis(PFGE)indicatedthatL.plantarumwassuccessful inassertingitselfasapredominantstrain,whichdidnot hap-penwithL.fermentum.

L.plantarumshowedbetterperformanceasastarterculture forsourcassavastarchproductioninapilot-scale fermenta-tionprocess.Cassavastarchproducedbythisstrainwasfound tohavethehighestvalueofTTAwhencomparedwiththe otherstartercultures;the pHandexpansioncapacitywere not different from the valuesobtained by the commercial product;also,therewasnoevidenceofpathogens.Therefore, furtherstudiesarerequiredfortheestablishmentofastarter culturethatwillcontributetothestandardizationofthe cas-savafermentationconditions,therebyensuringhigherquality productsandconsumeracceptability.

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

r

e

f

e

r

e

n

c

e

s

1.HowelerR,LutaladioN,ThomasG.TitleofSubordinate Document.SaveandGrow:Cassava:AGuidetoSustainable ProductionIntensification.Rome:FoodandAgriculture OrganizationoftheUnitedNations;2013.

http://www.fao.org/docrep/018/i3278e/i3278e.pdf.

2.IBGE,ftp://ftp.ibge.gov.br/ProducaoAgricola/

Levantamento SistematicodaProducaoAgricola

[mensal/Comentarios/lspa201403comentarios.pdf]Titleof

subordinatedocument.Indicadores:produc¸ãoagrícola: levantamentosistemáticodaproduc¸ãoagrícola:marc¸ode2014;

2014.

3.CookeRD,CourseyDG.Cassava:AMajorCyanide-Containing

FoodCrop.CyanideinBiology.NewYork:AcademicPress; 1981:93–114.

4.NoutMJR,SarkarPK.Lacticacidfoodfermentationin

tropicalclimates.AntonievanLeeuwenhoek.

1999;76(1–4):395–401.

5.FranzCMAP,HuchM,MatharaJM,etal.Africanfermented

foodsandprobiotics.IntJFoodMicrobiol.2014;190:84–96.

6.LacerdaICA,MirandaRL,BorelliBM,etal.Lacticacid

bacteriaandyeastsassociatedwithspontaneous

fermentationsduringtheproductionofsourcassavastarch

inBrazil.IntJFoodMicrobiol.2005;105(2):213–219.

7.CoulinP,FarahZ,AssanvoJ,SpillmannH,PuhanZ.

Characterisationofthemicrofloraofattiéké,afermented

cassavaproduct,duringtraditionalsmall-scalepreparation.

IntJFoodMicrobiol.2006;106(2):131–136.

8.CrispimSM,NascimentoAMA,CostaPS,etal.Molecular

identificationofLactobacillusspp.associatedwithpuba,a

Brazilianfermentedcassavafood.BrazJMicrobiol.

2013;44(1):15–21.

9.KostinekM,SpechtI,EdwardVA,etal.Characterisationand

biochemicalpropertiesofpredominantlacticacidbacteria

fromfermentingcassavaforselectionasstartercultures.Int

JFoodMicrobiol.2007;114(3):342–351.

10.WilfridPadonouS,NielsenDS,HounhouiganJD,ThorsenL,

NagoMC,JakobsenM.ThemicrobiotaofLafun,anAfrican

traditionalcassavafoodproduct.IntJFoodMicrobiol.

2009;133(1–2):22–30.

11.LacerdaICA,GomesFCO,BorelliBM,etal.Identificationof

thebacterialcommunityresponsiblefortraditional

fermentationduringsourcassavastarch,cachac¸aandminas

cheeseproductionusingculture-independent16srRNAgene

sequenceanalysis.BrazJMicrobiol.2011;42:650–657.

12.AidooKE,NoutMJ,SarkarPK.Occurrenceandfunctionof

yeastsinAsianindigenousfermentedfoods.FEMSYeastRes.

2006;6(1):30–39.

13.CapliceE,FitzgeraldGF.Foodfermentations:roleof

microorganismsinfoodproductionandpreservation.IntJ

FoodMicrobiol.1999;50(1):131–149.

14.AtrihA,RekhifN,MoirAJG,LebrihiA,LefebvreG.Modeof

action,purificationandaminoacidsequenceofplantaricin

C19,ananti-ListeriabacteriocinproducedbyLactobacillus

plantarumC19.IntJFoodMicrobiol.2001;68(1):93–104.

15.MessensW,DeVuystL.Inhibitorysubstancesproducedby

Lactobacilliisolatedfromsourdoughs—areview.IntJFood

Microbiol.2002;72(1–2):31–43.

16.MiduraTF,BryantRG.Samplingplans,samplecollection,

shipment,andprepartionsforanalysis.In:DownesFP,ItoK,

eds.CompendiumofMethodsfortheMicrobiologicalExamination ofFoods.4thed.Washington:AmericanPublicHealth

Association(APHA);2001:13–23.Chapter2.

17.NorrisJR,BerkeleyRCW,LoganNA,O’donnellAG.Thegenera

BacillusandSporolactobacillus.In:StarrMP,StolpH,Truper

onHabitats,IsolationandIdentificationofBacteria.

1981:1711–1742.

18.HoffmanCS,WinstonF.Aten-minuteDNApreparationfrom

yeastefficientlyreleasesautonomousplasmidsfor

transformaionofEscherichiacoli.Gene.1987;57(2–3):267–272.

19.LaneDJ.16S/23SrRNAsequencing.In:StackebrandtE,

GoodfellowM,eds.NucleicAcidTechniquesinBacterial

Systematics.NewYork:Wiley;1991:115–175.

20.BrightwellG,BoeremaJ,MillsJ,MowatE,PulfordD.

Identifyingthebacterialcommunityonthesurfaceof

IntraloxTM_{beltinginameatboningroomby}

culture-dependentandculture-independent16SrDNA

sequenceanalysis.IntJFoodMicrobiol.2006;109(1):47–53.

21.TamuraK,StecherG,PetersonD,FilipskiA,KumarS.MEGA6:

molecularevolutionarygeneticsanalysisversion6.0.MolBiol

Evol.2013;30(12):2725–2729.

22.TorrianiS,FelisGE,DellaglioF.DifferentiationofLactobacillus

plantarum,L.pentosus,andL.paraplantarumbyrecAgene

sequenceanalysisandmultiplexPCRassaywithrecA

gene-derivedprimers.ApplEnvironMicrobiol.

2001;67:3450–3454.

23.daSilva-FilhoEA,DosSantosSKB,doMonteResendeA,de

MoraisJOF,deMoraisMAJr,SimõesDA.Yeastpopulation

dynamicsofindustrialfuel-ethanolfermentationprocess

assessedbyPCR-fingerprinting.AntonievanLeeuwenhoek.

2005;88(1):13–23.

24.LieckfeldtE,MeyerW,BörnerT.Rapididentificationand

differentiationofyeastsbyDNAandPCRfingerprinting.J

BasicMicrob.1993;33(6):413–425.

25.LachanceM-A,BowlesJM,StarmerWT,BarkerJSF.

KodamaeakakaduensisandCandidatolerans,twonew

ascomycetousyeastspeciesfromAustralianHibiscus

flowers.CanJMicrobiol.1999;45(2):172–177.

26.OfficialMethodsofAnalysisofAOACINTERNATIONAL.20thEd. Gaithersburg,MD,USA:AOACINTERNATIONAL;2016.

www.eoma.aoac.org.

27.BennettRW,BelayN.Bacilluscereus.In:DownesFP,ItoK,

eds.CompendiumofMethodsfortheMicrobiologicalExamination ofFoods.Washington,DC:AmericanPublicHealth

Association;2001:311–316.

28.SanniAI,Morlon-GuyotJ,GuyotJP.Newefficient

amylase-producingstrainsofLactobacillusplantarumandL.

fermentumisolatedfromdifferentNigeriantraditional

fermentedfoods.IntJFoodMicrobiol.2002;72(1–2):53–62.

29.YarrowD.Methodsfortheisolation,maintenance

classificationandidentificationofyeasts.In:KurtzmanCP,

FellJW,eds.TheYeasts:ATaxonomicStudy.Amsterdam:

Elsevier;1998:77–100.

30.MaedaKC,CeredaMP.Avaliac¸ãodeduasmetodologiasde

expansãoaofornodopolvilhoazedo.FoodSciTechnol.

2001;21(2):139–143.

31.Brazil.MinistériodaSaúde.Resoluc¸ãoRDCn◦12,de02de

janeirode2001.Aprovaoregulamentotécnicosobreospadrões microbiológicosparaalimentos.DiárioOficialdaUnião,sec¸ão1;

2001.

32.Brazil.Decreton◦12.486,de20deoutubrode1978.Aprova

normastécnicasespeciaisrelativasaalimentosebebidas.Diário

OficialdoEstadodeSãoPaulo;1978:3–25.

33.KostinekM,SpechtI,EdwardVA,etal.Diversityand

technologicalpropertiesofpredominantlacticacidbacteria

fromfermentedcassavausedforthepreparationofGari,a

traditionalAfricanfood.SystApplMicrobiol.

2005;28(6):527–540.

34.ObilieEM,Tano-DebrahK,Amoa-AwuaWK.Souringand

breakdownofcyanogenicglucosidesduringtheprocessing

ofcassavaintoakyeke.IntJFoodMicrobiol.2004;93(1):115–121.

35.OguntoyinboFA,DoddCER.Bacterialdynamicsduringthe

spontaneousfermentationofcassavadoughingari

production.FoodControl.2010;21(3):306–312.

36.KimaryoVM,MassaweGA,OlasupoNA,HolzapfelWH.The

useofastartercultureinthefermentationofcassavaforthe

productionof“kivunde”,atraditionalTanzanianfood

product.IntJFoodMicrobiol.2000;56(2–3):179–190.

37.OmarNB,AmpeF,RaimbaultM,GuyotJ-P,TailliezP.

MoleculardiversityoflacticacidbacteriafromCassavasour

starch(Colombia).SystApplMicrobiol.2000;23(2):285–291.

38.OguntoyinboFA.EvaluationofdiversityofCandidaspecies

isolatedfromfermentedcassavaduringtraditionalsmall

scalegariproductioninNigeria.FoodControl.

2008;19(5):465–469.

39.OyewoleOB.Characteristicsandsignificanceofyeasts’

involvementincassavafermentationfor‘fufu’production.

IntJFoodMicrobiol.2001;65(3):213–218.

40.NoutMJR.Ecologyofacceleratednaturallacticfermentation

ofsorghum-basedinfantfoodformulas.IntJFoodMicrobiol.

1991;12(2):217–224.

41.BoonnopK,WanapatM,NontasoN,WanapatS.Enriching

nutritivevalueofcassavarootbyyeastfermentation.Sci

Agric.2009;66(5):629–633.

42.VogelmannSA,SeitterM,SingerU,BrandtMJ,HertelC.

Adaptabilityoflacticacidbacteriaandyeaststosourdoughs

preparedfromcereals,pseudocerealsandcassavaanduseof

competitivestrainsasstarters.IntJFoodMicrobiol.

2009;130(3):205–212.

43.GálvezA,AbriouelH,LópezRL,OmarNB.Bacteriocin-based

strategiesforfoodbiopreservation.IntJFoodMicrobiol.

2007;120(1–2):51–70.

44.NardiRMD,SantoroMM,OliveiraJS,etal.Purificationand

molecularcharacterizationofantibacterialcompounds

producedbyLactobacillusmurinusstrainL1.JApplMicrobiol.

2005;99(3):649–656.

45.BealesN.Adaptationofmicroorganismstocold

temperatures,weakacidpreservatives,lowpH,andosmotic

stress:areview.ComprRevFoodSciFoodSaf.2004;3(1):1–20.

46.ManteES,Sakyi-DawsonE,Amoa-AwuaWK.Antimicrobial

interactionsofmicrobialspeciesinvolvedinthe

fermentationofcassavadoughintoagbelimawithparticular

referencetotheinhibitoryeffectoflacticacidbacteriaon

entericpathogens.IntJFoodMicrobiol.2003;89(1):41–50.

47.MugulaJK,NarvhusJA,SørhaugT.Useofstarterculturesof

lacticacidbacteriaandyeastsinthepreparationoftogwa,a

Tanzanianfermentedfood.IntJFoodMicrobiol.

2003;83(3):307–318.

48.ObadinaAO,OyewoleOB,SanniLO,TomlinsKI.

Bio-preservativeactivitiesofLactobacillusplantarumstrainsin

fermentingCasssava‘fufu’.AfrJBiotechnol.2006;5:620–623.

49.HuchM,HanakA,SpechtI,etal.UseofLactobacillusstrains

tostartcassavafermentationsforGariproduction.IntJFood

Microbiol.2008;128(2):258–267.

50.EdwardVA,HuchM,DortuC,etal.Biomassproductionand

small-scaletestingoffreeze-driedlacticacidbacteriastarter

strainsforcassavafermentations.FoodControl.